Conceptual Dependency KR Chowdhary, Professor, Department of - - PowerPoint PPT Presentation

• Dept. of CSE, MBM Engg.

Conceptual Dependency

KR Chowdhary, Professor, Department of Computer Science & Engineering, MBM Engineering College, JNV University, Jodhpur,

• Dept. of CSE, MBM Engg.

KNOWLEDGE REPRESENTATION:

1970’S: Conceptual Dependency Grammar

A number of authors in AI have addressed the question of the 'concept'-based organisation of knowledge and we use two examples to illustrate this.

• We consider a verb-oriented organisation of

knowledge proposed by Schank: Conceptual Dependency Grammar.

• Dept. of CSE, MBM Engg.

KNOWLEDGE REPRESENTATION:

1970’S: Conceptual Dependency Grammar

Conceptual dependency (or CD) is a theory of how to represent the meaning of natural language sentences in a way that:

• First, facilitates for drawing inferences from the

sentences.

• Second, it has been argued that the representation

(CD) is independent of the language in which the sentences were originally stated

• Dept. of CSE, MBM Engg.

KNOWLEDGE REPRESENTATION:

1970’S: Conceptual Dependency Grammar

• Schank's (1975) Conceptual Dependency Theory was developed as

a part of a natural language comprehension project.

• Schank's claim was that sentences can be translated into basic

concepts expressed as a small set of semantic primitives.

• Conceptual dependency allows these primitives, which signify

meanings, to be combined to represent more complex meanings.

• Schank calls the meaning propositions underlying language

"conceptualisations".

• Dept. of CSE, MBM Engg.

Roger Schank’s project is the ‘representation of meaning in an unambiguous language-free manner’ (1973). ‘Any two utterances that can be said to mean the same thing, whether they are in the same or different languages, should be characterised in only one way by the conceptual structure’ (1973) (CG)

Three elemental kinds of concepts: a nominal and an action together with their modifiers

• Dept. of CSE, MBM Engg.

Three elemental kinds of concepts:

• 1. Concept

a nominal

• an abstract or concrete object that invokes an image;
• cars are concrete objects;
• gravity is an abstract concept;
• a nominal produces a picture (PP)

     

be can

• Dept. of CSE, MBM Engg.
• 2. Concept

an action

• what a nominal does?
• something an animate nominal does to an object;
• there are primitive ACTions and derived ACTions;

     

be can

• Dept. of CSE, MBM Engg.
• 3. Concept

a modifier

• a modifier modifies a nominal or an action;
• a modifier specifies an action or a nominal;

     

be can

• Dept. of CSE, MBM Engg.

3a Concept a picture modifier or aider (PA)

• a

car 3b Concept an action modifier or aider (AA)

• he

run

     

be can

     

be can

PA blue

AA quickly

• Dept. of CSE, MBM Engg.

CD theorists argue that 'the CD representation of a sentence is built not

• ut of primitives corresponding to the words used in the sentence, but

rather out of conceptual primitives that can be combined to form the meanings of words in any particular language'

Diagrammatic conventions Conceptual Tenses Primitive conceptual categories Primitive Acts CD Building Blocks

• Dept. of CSE, MBM Engg.

Conceptual Categories

Three elemental kinds of concepts – conceptual categories (PP, ACT, PA and AA) – relate to each other in specified ways. These relations are called dependencies by Schank.

• Dept. of CSE, MBM Engg.

Conceptual Categories

In a dependency relation, one partner or item is dependent and the

• ther dominant or governing; a governordependent is a partially
• rdered relationship
• A dependent must have a governor and is understood in terms
• f the governor
• A governor may or may not have dependent(s) and has an

independent existence

• A governor can be a dependent

PP and ACT are inherently governing categories, PA and AA are inherently dependent.

• Dept. of CSE, MBM Engg.

Conceptual Categories

For a conceptualisation to exist, there must be at least two governors:

Sally stroked her fat cat

‘Sally’ name of an object (nominal  PP) ‘stroked’ name of an action (ACT)

PP: Sally, cat, her [Sally] ACT: to stroke PA: fat Governors: Sally, stroke, cat Dependent: PP (cat) on ACT (stroke) PA (fat) on PP (cat) PP (cat) on PP (her[Sally])

• Dept. of CSE, MBM Engg.

Conceptual Categories

Sally stroked her fat cat

1. Sally and stroking are necessary for conceptualisation: there is a two-way dependency between each other: Sally  stroke 2. Sally’s cat cannot be conceptualised without the ACT stroke

 it has an objective dependency on stroke

Sally  stroke cat.

 O

• Dept. of CSE, MBM Engg.

Conceptual Categories

Sally stroked her fat cat

3. The concept ‘cat’ is the governor for the modifier ‘fat’: Sally  stroke cat



fat

 O

• Dept. of CSE, MBM Engg.

Conceptual Categories

Sally stroked her fat cat

4. The concept PP(cat) is also governed by the concept PP(Sally) through a prepositional dependency: Sally  stroke cat



POSS-BY fat Sally[her]

A Conceptual Dependency Network

 O

• Dept. of CSE, MBM Engg.

Conceptual Cases

Dependents that are required by an ACT are called Conceptual Cases: Consider the following sentences: Objective Case (O) 1. John took the book PP[John]  ACT[took] PP[book] Recipient Case (R) 2. John took the book from Mary PP[John] R PP



ACT < PP[Mary]

 O

• PP

• Dept. of CSE, MBM Engg.

Conceptual Cases

Instrumental Case (I) 3. John ate the ice cream with a spoon PP[John] PP[John]  ACT [eat]



do

• PP[spoon]

 I

PP[ice cream]

• Dept. of CSE, MBM Engg.

Conceptual Cases

Directive Case Relation (D) 4. John drove his car to London from Guildford John  do London car  drove

 POSS-BY

< Guildford John ( ) indicates causality

 

D

• Dept. of CSE, MBM Engg.

Prepositional Dependency

Consider the following sentences: Possession

• This is Sally’s cat  one PP (Sally) provides specific information

about another PP (cat) - Cat

 POSS-BY

Sally Location

• Sally is in London -

London

 LOC

Sally Containment

• The glass contains water -

Water

 CONT

Glass

• Dept. of CSE, MBM Engg.

Underlying ACTs

Giving and Taking OR just TRANSferring

I took a book from Sally

PP[I] R PP [I]  ATRANS < PP[Sally]

• PP [book]

• Dept. of CSE, MBM Engg.

Sentence = ‘I took a book from Sally’

Arrows indicate the direction of dependency Double arrow indicates two way link between actor and action Diagrammatic Conventions

• indicates object case relation; R

indicates recipient case relation Conceptual Tenses Objects (Picture Producers: PP): Sally, I , book Primitive conceptual Categories ATRANS indicates transfer (of possession) Primitive Acts

Elaboration CD Building Block

• Dept. of CSE, MBM Engg.

Underlying ACTions

Conceptualise or think about an idea (think)

CONC

Action of attending or focusing a sense organ towards a stimulus (listen)

ATTEND

Actions of producing sounds (say)

SPEAK

Construction by an animal of new information of old information (decide)

MBUI LD

Transfer of mental information between animals or within an animal (tell)

MTRANS

Expulsion of an object from the object of an animal into the physical world (cry)

EXPEL

Taking in of an object by an animal to the inside of that animal (eat)

I NGEST

Grasping of an object by an actor (grasp)

GRASP

Movement of a body part of an animal by that animal (kick)

MOVE

Application of a physical force to an object (push)

PROPEL

Transfer of the physical location of an object (go)

PTRANS

Transfer of an abstract relationship such as possession ownership or control (give)

ATRANS Elaboration Primitive Act

• Dept. of CSE, MBM Engg.

Underlying ACTs

I gave a book to Sally

Sally R I  gave < I PP[Sally] R PP [I]  ATRANS < PP[I]

• book
• PP [book]

• Dept. of CSE, MBM Engg.

Underlying ACTs – The TRANSfer ACT

• Transfer of an abstract relationship: possession
• f property  give, take; Abstract TRANSfer

(ATRANS)

• Transfer of the physical location of an object 

go, come; Physical TRANSfer (PTRANS)

• Transfer of mental information  ask, tell;

Mental TRANSfer (MTRANS)

• Dept. of CSE, MBM Engg.

Underlying ACTs – STATI CS and DYNAMI CS

• PROPELling an object by applying physical

force  push, pull; PROPEL

• MOVEment of a body part by its owner 

kick; MOVE

• GRASPing of an object by an actor 

catching, clutching; GRASP

• Dept. of CSE, MBM Engg.

Underlying ACTs – Sustaining life; processing things

• INGESTing of an object by an animal

 eat, drink; INGEST

• EXPELling of something from a body
• f an animal  crying [tears]

• Dept. of CSE, MBM Engg.

Underlying ACTs – Stimulus and Response

• Mentally BUILDing new information out of
• ld  deciding, inferring; MBUILD
• SPEAKing by producing sounds  talking;

SPEAK

• ATTENDing to a stimulus by focusing a sense
• rgan towards it  seeing, hearing; ATTEND

• Dept. of CSE, MBM Engg.

Conceptual Tenses

Any conceptualisation can be modified as a whole by a conceptual tense. John took the book – John  took – can be denoted by looking at the lemma take (from which the past tense took was derived): p(ast) John  take

• Dept. of CSE, MBM Engg.

Conceptual Tenses

Conditional c Timeless Delta Present nil Negative / Interrogative ? Continuing k Finished Transition tf Start Transition ts Transition t Future f Past p Elaboration Symbol

• Dept. of CSE, MBM Engg.

Conceptual Tenses

• John will be taking the book can be

described as: f(uture) John  taking or John  take

• John is taking the book can be described as:

k John  taking or John  take

• Dept. of CSE, MBM Engg.

Semantic Nets and Conceptual Dependency Representation: A comparison

• Semantic Nets only provide a structure into

which nodes representing information can be placed

• Conceptual Dependency representation, on the
• ther hand, provides both a structure and a

specific set of primitives out of which representations of particular pieces of information can be constructed

• Dept. of CSE, MBM Engg.

Advantages of using Conceptual Dependency Grammar for representing knowledge and reasoning with a CD knowledge base

• 1. The organisation of knowledge in terms of the primitives

(or 'primitive acts') leads to a fewer inference rules.

• 2. Many inferences are already contained in the

representation itself.

• 3. The initial structure that is built to represent the

information contained in one sentence will have holes in it that have to be filled in: holes(e.g. book, john, etc.) which will serve as attention focusers for subsequent sentences.

• Dept. of CSE, MBM Engg.

Disadvantages of using Conceptual Dependency Grammar for representing knowledge and reasoning with a CD knowledge base

• 1. CD requires all knowledge to be broken down

into 11 primitives: sometimes inefficient and sometimes impossible

• 2. CD is essentially a theory of the representation of

events: though it is possible to have an event- centred view of knowledge but not a practical proposition for storing and retrieving knowledge

• Dept. of CSE, MBM Engg.

Disadvantages of using Conceptual Dependency Grammar for representing knowledge and reasoning with a CD knowledge base

• 3. May be difficult or impossible to design a

program that will reduce sentences to canonical

• form. (Provably not possible for monoids, which

are simpler than natural language).

• 4. Computationally expensive to reduce all sentences

to the primitives.